Zhicheng Han and Renata Forste, Department of Sociology

Based on a nationally representative sample of women in Nepal aged 15 to 49, this study examines the association between domestic violence and women’s empowerment at both the individual and community level. In a sample of 3,349 ever married woman in Nepal, I use logistic regression to estimate the models. I find that female autonomy and residence in the Terai region are associated with higher odds of experiencing domestic violence. In addition, interactions indicate that female autonomy lowers the odds of domestic violence across ecological regions, religion, ethnicity and residency.

Domestic violence or intimate partner violence (IPV) is prevalent, especially in less developed countries. In Nepal, over twenty percent of women reported having experienced physical violence, and over ninety percent of that physical violence was committed by a current or former husband. About fifteen percent of physical violence experienced by women in the past twelve months resulted in deep wounds, broken bones, broken teeth, or other serious injuries (NDHS, 2011). Cultural factors within the community often function independently of socio-economic conditions or women’s empowerment on individual levels. Research in Bangladesh show that individual-level women’s status can elevate the risk of domestic violence, while community-level female status significantly lowers the risk of violence (Koenig et al., 2003). I model women’s empowerment at the individual level to see if it is associated with domestic violence; in addition, I examine the effect of ecological region, religion, caste position, and urban residence on odds of domestic violence independently. Then, I run interactions between factors that impact odds of domestic violence and women’s autonomy. Other factors that affect the occurrence and reporting of IPV that controlled for are: wealth, level of education, employment, women’s age, age at first marriage, and number of living children. I examine factors predictive of women experiencing domestic violence and I expect that the more a culture empowers women, the less likely they are to experience IPV. Women in regions or cultures that do not empower woman might increase their odds of experiencing domestic violence.

Contrary to the literature, being a member of the higher caste or living in urban areas is associated with increased odds of IPV when social economic factors are controlled for (Model 3). Region wise, the terai region shows significantly higher odds of domestic violence than the hill or mountain regions. The odds of Christians and Muslims ever experiencing domestic violence are about 80% higher than for other religions. The result is only significant when social economic factors are controlled for (Model 3). Among social economic factors included in other studies, the number of living children did not have a significant impact of the odds of ever experiencing domestic violence. Interactions between community variables and empowerment showed no significant result. Empowerment reduces the odds of domestic violence regardless if the respondents have urban or rural residence, Christian/Muslim religions or other, residence in Terai region or are members of the higher caste.

A number of variables considered stressors were associated with increased odds of domestic violence. Wealth and women’s age were correlated with increased odds of domestic violence. However other stressors do not have a significant relationship with domestic violence. Number of living children dose not significantly impact the odds of women experiencing domestic violence when other socio-economic factors are controlled for, suggesting that other theories might be needed to explain odds of domestic violence. Most of the factors focused on by stressor theory are on either family or individual level according to Carlson’s model, it ignored the social-structural factors that include larger cultural framework and power dynamic outside of the family unit (Carlson, 1984). My findings are however, in accordance with others who included these factors such as Narved’s 1994 study in Bangladesh. Stressor theory is but a part of the complicated landscape of IPV and its significant might subside when social and cultural context is applied.

My main finding is contrary to the Narved’s worry that empowerment might increase domestic violence in culturally conservative areas and consequent research (Koenig et.al, 2006), I found no evidence that increased female autonomy increased the odds of domestic violence in areas that do not culturally empower women. I found that even in Treai region where odds of domestic violence are higher than all other regions, higher level of female autonomy is associated with lowered odds of domestic violence. In urban areas where odds of domestic violence are higher, autonomy is still significant. My results suggest that women’s empowerment has a real impact on domestic violence. Woman who have more say in household matters learnt to stand up for themselves and assert more power in family dynamics. Previous studies theorized the that on family level, shifting in power induces conflict because husbands feel their power is being stripped away (Carlson,1984). Empowering women might indeed cause husbands to feel insecure and increase domestic violence. But only when the shift in power is nascent do husbands feel threatened; therefore, the negative effect of empowerment women is likely transitional. Long term observations will show positive effects of giving women say in household matters.

I also found that being a member of higher caste increased the odds of domestic violence. This further proves that cultural factors influence domestic violence. Being a member of the higher caste means more rigid social expectations for one to behave according to the traditional family roles. Members of higher caste might have held a less fluid conception of social norms and felt more threatened when the norms were challenged. This could also be explained by the theory that husbands use violence to control their partners’ behavior when they have no other resources, as women in higher castes might be more socially independent. They are more likely to sever the social ties that kept previous generations of women’s behavior in check. This theory can also explain why when wealth is controlled for, urban residence significantly increases the odds of domestic violence. In an urban community with weaker social ties, husband might lack other social means to control their wives’ behavior so they resort to violence. I found the increase of domestic violence in religious groups that has weaker social ties as well. Christian and Muslim religion increased odds of domestic violence because these groups are often more recent immigrants from India and other South Asian countries with weaker communities and social connections in Nepal. This theory requires future research with more information on husbands and community. Variables such as husband’s education, employment, income and participation in local community should be sufficient to examine if husband’s socio-economic standings effect the odds of domestic violence.

Giving women more resources might lead to short term negative consequences. But giving women more say in household and social matters will lead to a change to the norm that benefits all. Women would learn to stand up for themselves and ask for a more equal society. Polices and developmental projects should be designed to give women more say in politics, economics and social matters. Grass root movements working towards empowering women might be especially effective as one can expect less residence to normative changes in gender roles. Strengthening communities around women in urban areas can foster a social norm against domestic violence. Empowering women has a pivotal role to play in the fight against domestic violence.